Adversarial AI

prompt injection defense, adversarial AI course, LLM security, AI red teaming, jailbreak LLM, context poisoning, RAG poisoning, PyRIT tutorial, indirect prompt injection, LLM application security

advanced

en

• Patch prompt injection with layered defense-in-depth before shipping an LLM feature
• Defend the agent attack surface against RAG poisoning, document payloads, memory poisoning, and tool-output hijacking
• Stand up automated jailbreak tooling with DeepTeam, PyRIT, and Mindgard
• Apply the PAIR, TAP, and GCG attack algorithms and the BYO-attacker pattern
• Cache attack trajectories to turn expensive discovery into cheap CI security gates
• Run an internal red-team program with attack taxonomy, finding lifecycle, runtime monitoring, and alignment regression

• Url: https://aiacademy.anthropos.work/chapters/adversarial-ai-intro/ · Adversarial AI: Start Here · Position: 1 · Orientation across the four-chapter Adversarial AI skill path — defender → threat-surface → offense → program — covering everything from your first prompt-injection patch to running an internal red-team practice
• Url: https://aiacademy.anthropos.work/chapters/prompt-injection-defense/ · Prompt Injection Defense Foundations · Position: 2 · Why prompt injection exists, how attackers exploit it, and the layered defense every LLM-powered feature needs before shipping
• Url: https://aiacademy.anthropos.work/chapters/context-poisoning-indirect-injection/ · Context Poisoning & Indirect Injection · Position: 3 · Map and defend the agent-era attack surface — RAG poisoning, document-borne payloads, memory poisoning, and tool-output hijacking that direct-injection defenses don't reach
• Url: https://aiacademy.anthropos.work/chapters/automated-jailbreak-tooling/ · Automated Jailbreak Tooling · Position: 4 · Stand up automated offensive tooling for AI red-teaming — DeepTeam, PyRIT, Mindgard; attacker LRMs and BYO-attacker pattern; PAIR, TAP, GCG algorithms; and trajectory caching that turns expensive discovery into cheap CI gates
• Url: https://aiacademy.anthropos.work/chapters/ai-red-teaming/ · AI Red Teaming & Adversarial Evaluation · Position: 5 · Run a red-team program for production AI — taxonomy, the finding lifecycle, runtime monitoring, alignment regression, and the internal playbook

• Patch prompt injection with layered defense-in-depth before shipping an LLM feature
• Defend the agent attack surface against RAG poisoning, document payloads, memory poisoning, and tool-output hijacking
• Stand up automated jailbreak tooling with DeepTeam, PyRIT, and Mindgard
• Apply the PAIR, TAP, and GCG attack algorithms and the BYO-attacker pattern
• Cache attack trajectories to turn expensive discovery into cheap CI security gates
• Run an internal red-team program with attack taxonomy, finding lifecycle, runtime monitoring, and alignment regression

Course Mode: online · Course Workload: PT100M · Mode: online